verl-rl-training
仓库创建 2025年11月3日最近提交 1 个月前SkillHot 收录 21 天前
▸ 精选理由
针对生产级 RLHF 与大规模训练优化,适合有集群需求的研发团队。
▸ 风险提示
常需企业级后端与大规模算力,部署复杂且依赖私有基础设施。
这个 Skill 做什么
verl 提供面向大规模 LLM 的高效 RL 训练与混合流框架实现。
做大规模 LLM 的强化学习(如 RLHF、GRPO、PPO)后训练时就会用到。verl 提供了生产级的训练框架(HybridFlow),能在多种基础设施后端上高效跑分布式 RL 训练和回放流水线。适合需要可扩展、低开销训练流程、或想把研究级算法投入生产的场景。特点是性能与工程化并重,方便迁移到实际集群。
▸ 展开 SKILL.md 英文原文
Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.
1.1w
Stars
817
Forks
40
仓库内 Skill
+704
7 日增星
安装 / 使用
给你的 Agent 一句话(通用)
帮我安装这个 skill:https://raw.githubusercontent.com/Orchestra-Research/AI-Research-SKILLs/main/06-post-training/verl/SKILL.md或 curl 直取 SKILL.md
curl -fsSL "https://raw.githubusercontent.com/Orchestra-Research/AI-Research-SKILLs/main/06-post-training/verl/SKILL.md"SKILL.MD 节选查看完整文件 ↗
# verl: Volcano Engine Reinforcement Learning for LLMs verl is a flexible, efficient, and production-ready RL training library for large language models from ByteDance's Seed team. It implements the HybridFlow framework (EuroSys 2025) and powers models like Doubao-1.5-pro achieving O1-level performance on math benchmarks. ## When to Use verl **Choose verl when you need:** - Production-ready RL training at scale (tested up to 671B parameters) - Flexibility to swap backends (FSDP ↔ Megatron-LM ↔ vLLM ↔ SGLang) - Support for multiple RL algorithms (PPO, GRPO, RLOO, REINFORCE++, DAPO) - Multi-turn rollout with tool calling for agentic workflows - Vision-language model RL training **Consider
via SKILL·HOT · 数据来自 GitHub 公开信息 · 原文版权归作者所有